Open-Source MRI Sequence Development: Simulation to Scanner

Overview

This interdisciplinary mini-course introduces an end-to-end workflow for MRI sequence development using open-source tools. Students will design pulse sequences in Pulseq, simulate and validate them with KomaMRI, and reconstruct simulated or acquired data using BART and MRIReco. Hands-on laboratory sessions connect simulation to scanner experiments through Pulseq interpreters and OpenRecon, with an emphasis on reproducible workflows and scientific interpretation.

There will be an emphasis on research rigor, reproducibility and transparency, by using open-source tools and the creation of reproducible pipelines. Intended for students with basic programming experience and an interest in MRI and computational methods.

Learning goals

By the end of this mini-course, students will be able to:

Design basic MRI pulse sequences using Pulseq.
Simulate MRI experiments using Bloch simulations in KomaMRI.
Reconstruct MRI data using BartIO and MRIReco.
Analyze the impact of sequence parameters on image contrast and artifacts.
Implement and iterate on sequence designs using a rapid prototyping workflow.
Evaluate sequence performance through simulation and scanner experiments.

Schedule & materials

Last updated: October 21, 2026

Scroll sideways to see all days.

Week	Monday	Wednesday	Friday
Week 1	Day 1 · Nov 9 Lecture: Open source, AI agents, and MRI simulations Coming soon Laboratory: Run and interpret a first Bloch simulation Coming soon	Day 2 · Nov 11 Lecture: RF excitation and selective pulses Coming soon Laboratory: Simulate the 2D RF excitation sequence Coming soon	Day 3 · Nov 13 Lecture: RF pulse validation Coming soon Scanner laboratory: Test your chosen 2D RF pattern; compare with simulations Coming soon
Week 2	Day 4 · Nov 16 Lecture: Gradients and Cartesian k-space Coming soon Laboratory: Simulate gradient moments and EPI artifacts Coming soon	Day 5 · Nov 18 Lecture: Radial and spiral trajectories Coming soon Laboratory: Compare undersampling, delays, and off-resonance Coming soon	Day 6 · Nov 20 Lecture: Trajectories and hardware limits Coming soon Scanner laboratory: Acquire and reconstruct radial data Coming soon
Week 3	November 23-27 · Thanksgiving break
Week 4	Day 7 · Nov 30 Lecture: Gradient echo, spin echo, and steady states Coming soon Laboratory: Simulate spoiling, banding, and relaxation Coming soon	Day 8 · Dec 2 Lecture: Coil sensitivities, parallel imaging, and compressed sensing Coming soon Laboratory: Reconstruct data using BART and MRIReco Coming soon	Day 9 · Dec 4 Lecture: Pulseq and OpenRecon Coming soon Scanner laboratory: Evaluate an acquisition and reconstruction pipeline Coming soon

Classroom: CCSR4107, except days 1 and 3: M218/218A.

Lecture: 9:30-10:20 a.m.
Lab: 10:30-11:50 a.m.

Friday scanning: After the lecture, we’ll head to Cima.X (3T3) in the Lucas Center.

Before the first class

Complete these steps: yourself; coding agent

Install and sign in to a desktop coding agent (not the CLI version), such as Codex or Claude Code.
Ask your coding agent to install Julia 1.12 and add Pluto to Julia’s global environment.
Ask your coding agent to clone the activities repository.
Open the cloned folder as a project in your coding agent to access AGENTS.md and the course skill.
Ask your coding agent to open Activity 0, then complete the activity to familiarize yourself with the environment and confirm:
- Correct environment setup, including Julia 1.12 (check VERSION).
- Pluto’s reactivity when inputs change.
- Use of the course-provided skill file by your coding agent.

Report any setup problems before the first class. This readiness check is ungraded.

Canvas

Enrolled students can access BIOS 214 on Canvas for announcements, submissions, grades, and any restricted material.